Human acute myelogenous leukemia (AML) is highly heterogeneous and, as demonstrated by this group, exhibits significant variability in sphingoiipid metabolism. Current therapy of AML is highly toxic, yielding ultimately inadequate outcomes for the vast majority of patients. Cell lines are limited in their representation of primary AML subtypes and manifest clonal evolution in culture, suggesting limitations in their relationship to the primary case hnaterial. The systematic study of these diseases thus requires access to primary samples representing a substantial pool of cases, and correlation of these samples to clinical data including treatment outcomes. Animal models of various kinds are available, but models representing direct leukemogenesis via expression of relevant oncogenic proteins and immunodeficient xenografts propagating primary AML, provide relevant platforms for studying developing therapies. In this Core, materials and models will be provided to each Project to allow completion of proposed objectives. The following specific aims will be pursued: Specific Aim 1: Expand and maintain the PSHCI leukemia cell bank to provide programmatic access to cell and plasma samples from a broad variety of AML patients and normal hematopoietic controls. Samples will be cryopreserved in multiple aliquots to allow repeated interrogation via developing technologies. Clinical data will be collected and available to provide biologically meaningful categorization of AMLs. Specific Aim 2: Develop and maintain a menu of animal models in which to test promising program-derived therapies. Two such models are available. Subaim 2A: Xenograft models of primary human leukemia have been developed in N0D/SCID/IL2rY